Storage has ceased to be a mere cost factor in the enterprise IT
infrastructure, and is viewed as a revenue enabler. The primary challenge before
solution partners and CIOs now, is to manage the entire storage architecture,
which, instead of having a small footprint, is varied and heterogeneous.
There is no doubt that information storage is given top priority amongst
enterprises, as it is a very potent tool for being competitive in today's
dynamic business environment. This has led to a tremendous explosion in demand
for storage infrastructure. While the importance of storage is undeniably
growing, enterprises are trying hard to transform storage from being viewed as a
cost component to the IT infrastructure, to a revenue generator.
The cost of storage technologies are declining daily. But the dominance of
fiber channel technology in the networked storage era has led to a situation
wherein the cost of manpower required to administer storage is escalating day by
day. And with shrinking IT budgets, storage administrators are finding it
increasingly difficult to meet the demands placed upon them while they try to
manage the storage infrastructure.
This has become a deterrent for enterprises in their efforts towards
transforming storage from a cost factor to a revenue enabler. Therefore, the
primary challenge before solution partners and CIOs is to manage the entire
storage architecture, which, instead of having a small footprint, has a varied
and heterogeneous footprint. This kind of a situation puts the performance of
business applications, and in turn the business, at risk.
WHY GO FOR SAM?
Storage Area Management (SAM) or consolidation of storage resources has
apparently evolved as a viable solution for addressing the above mentioned
challenges. However, not all SAM solutions address the problems faced by storage
administrators.
This is because, when considered from a business point of view SAM is all
about the quality of service. In a typical organization, storage administrators
have to manually manage the various elements of the storage infrastructure.
This may include file systems that are linked to the database, host bus
adopters, SAN fabric switches, the host server, the storage system itself and
the individual disk drive. The entire chain from drive to application has to be
linked. This is a labor intensive, error prone and slow process, which adds up
to delay and higher costs.
However, in today's scenario, businesses that are able to deliver high
levels of service easily will automatically reap benefits, whereas businesses
that cannot will be in trouble. For instance, when an application experiences
unexpected activity, it may not have the storage service that meets its
requirement. Consequently, performance and service deteriorates leading to
crashes and downtimes.
IMPLEMENTING SAM
While the implementation of a SAM practice seems the most obvious solution
in such a scenario, it is important for solution providers to recognize the
change and revisit the storage setup. It is because, if they have to effect
change, it has to be done not in months, but in a matter of hours.
They need to understand that the implementation of SAM should enable them to
manage the complex SAN infrastructures, storage resource management to address
process issues and storage provisioning to enable new applications to come
online quickly. In this manner, the SAM solution is able to deliver significant
TCO improvements and faster ROI.
However, in order to get the most out of a SAM implementation, it is
important to follow a layered model for storage management and the whole
industry is moving towards this. This type of a model relies on building layers
within the storage infrastructure and it is based on the industry wide standards
initiative - Common Information Model (CIM).
TO CIM OR NOT TO CIM
The CIM model enables management systems to recognize and exchange
information in a heterogeneous storage environment that houses different
solutions from various vendors. It offers a powerful foundation for
cross-platform SAM implementation.
Though the industry has not reached a consensus on the exact definition of
the layers, one model, which has identified four basic management layers is
slowly gaining prominence. The four layers comprise of reporting, path
management, event monitoring and automation control and they are addressed in
succession.
However all these four levels strongly depend upon the base level - discovery
- which serves as the building block for this model to work. It is because, at
the outset, organizations need to understand what storage resources they have
and where they are.
This is a very cumbersome process, which earlier used to comprise of reams of
spreadsheets, whiteboards and hand written notes. Ideally, a good SAM solution
should help administrators get a holistic perspective of all the elements
involved in a storage infrastructure right from the Logical Unit Number (LUN)
all the way to the application. And this has to be viewed from the application's
viewpoint.
This helps administrators to discover, track and locate the existing
resources and new resources that are being added into the storage
infrastructure. This apparently provides a very logical platform that can be
used to manage storage in a situation wherein enterprises prefer to build
storage infrastructures by sourcing different vendors.
BINDING ALL ELEMENTS TOGETHER
Once the solution provider discovers all the elements within the storage
system it is important for them to understand how the resources are being used.
This is because enterprises often resort to having buffer storage to meet
unforeseen circumstances, so that all mission-critical applications are
functional.
However, automated reporting and monitoring to determine what additional
resources are needed and when they are required comes out as a preferable
approach. From this viewpoint, the SAM solution should be able to optimize the
existing storage resources rather than rely on buffer storage capacity to keep
the applications up and running.
Hence, ideally, while addressing the reporting level in storage management,
the SAM solution should be able to provide an elemental overview across the
entire storage infrastructure so that they can continuously monitor storage from
LUN to the application. Further, complementing the comprehensive reporting and
monitoring functionality, the SAM solution should be able to provide
administrators with predictive management tools based on historic data so that
outages can be anticipated even before they happen and corrective actions can be
taken.
With such a reporting and monitoring system in place, partners can optimize
the existing usage resources and they can balance loads and shift excess
capacity from one server and application to another. This eliminates the need
for excess and "unnecessary" capacity, while existing
"excess" capacity and bandwidth are available elsewhere within the
existing infrastructure.
V Vivekanand is Sales
Director, SAM Solutions, APIA at Hitachi Data Systems
CHECKLIST FOR SAM IMPLEMENTATION
Model standard